# Free Will with a Little Bit of Science! (TM)(C)(R) **URL:** \ **Category:** Great Debates\ **Created:** [January 30, 2021, 5:53pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877 "2021-01-30T17:53:11Z")\ **Posts on this page:** 20\ **Page:** 9
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 1:59pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/161 "2021-02-17T13:59:26Z")
I didn’t answer your question. It’s easier to see in a Fuzzy Logic equivalent. It’s a center of gravity calculation. If 2+2 and 4+4 are membership functions then 3+3 lies between them. A playground teeter totter will balance the same way. It provides a generalized output without executing rules. ---
**Author:** ![Babale](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/babale/32/15666_2.png) [@Babale](https://boards.straightdope.com/u/Babale)\ **Post date:** [February 17, 2021, 2:30pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/162 "2021-02-17T14:30:05Z")
> [@Crane](#): > > BTW: I’m half way through the book you suggested _21 Lessons_ , very interesting discussion. Glad you are enjoying it! ---
**Author:** ![Babale](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/babale/32/15666_2.png) [@Babale](https://boards.straightdope.com/u/Babale)\ **Post date:** [February 17, 2021, 2:33pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/163 "2021-02-17T14:33:15Z")
> [@Crane](#): > > I didn’t answer your question. It’s easier to see in a Fuzzy Logic equivalent. It’s a center of gravity calculation. If 2+2 and 4+4 are membership functions then 3+3 lies between them. A playground teeter totter will balance the same way. It provides a generalized output without executing rules. I agree that there is no programming that tells the neural net exactly what to do in the event of 3+3, in your scenario. However, it still decides what to do by looking at the 2+2 and 4+4 situations and extrapolating from there. It does this by following internal rules, whether those were explicitly programmed or emergent as the result of training, doesn’t it? ---
**Author:** ![BeepKillBeep](https://avatars.discourse-cdn.com/v4/letter/b/9e8a1a/32.png) [@BeepKillBeep](https://boards.straightdope.com/u/BeepKillBeep)\ **Post date:** [February 17, 2021, 2:37pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/164 "2021-02-17T14:37:29Z")
This really isn’t correct as it relates to the research I’m doing. The resulting trained neural network, in your example, is a deterministic model of addition. There are other types of models besides deterministic and stochastic, but a generalized model is not one of them. Different types of models can be generalized. All generalization means is that the algorithm has been trained such that it functions properly for data that was not in the training data, and ideally, functions well for real-world data. ---
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 2:57pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/165 "2021-02-17T14:57:42Z")
> [@Babale](#): > > It does this by following internal rules That’s philosophical since the rules are a human way of describing how a thing operates. I’d say that generalization is ‘magic’ to the extent that it can respond to unlearned stimulae with outputs it has not experienced. The most significant non deterministic act of the brain is it’s ability to ask a question. A unique feature of the brain is to instantly know what it does not know: “Who was Rutherfords vice president?” Most likely you are able to immediately say ‘I don’t know’. A computer would have to do a memory search. So the brain is able to perceive that it does not know and ask questions to fill in the blanks. It can seek information. It can then generalize between the new categories of information. Obviously there is a mechanism to accomplish this. I believe that mechanism operates in the domain between deterministic and stochastic. ---
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 2:59pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/166 "2021-02-17T14:59:45Z")
> [@BeepKillBeep](#): > > All generalization means is that the algorithm has been trained such that it functions properly for data that was not in the training data Isn’t that the definition of volition? ---
**Author:** ![BeepKillBeep](https://avatars.discourse-cdn.com/v4/letter/b/9e8a1a/32.png) [@BeepKillBeep](https://boards.straightdope.com/u/BeepKillBeep)\ **Post date:** [February 17, 2021, 3:04pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/167 "2021-02-17T15:04:24Z")
I don’t see the connection. Can you explain how generalization relates to volition in your view? ---
**Author:** ![Babale](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/babale/32/15666_2.png) [@Babale](https://boards.straightdope.com/u/Babale)\ **Post date:** [February 17, 2021, 3:12pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/168 "2021-02-17T15:12:33Z")
> [@Crane](#): > > That’s philosophical since the rules are a human way of describing how a thing operates. I’d say that generalization is ‘magic’ to the extent that it can respond to unlearned stimulae with outputs it has not experienced. I don’t think it is - there are some excellent comp sci videos explaining how a neural network actually works, which I won’t embarass myself by trying to explain too in depth - but it does come back to simplifying the problem to yes/no states and making a guess based on those. Here is an example: [![](https://img.youtube.com/vi/BFdMrDOx_CM/maxresdefault.jpg "Inside a Neural Network - Computerphile") ](https://www.youtube.com/watch?v=BFdMrDOx_CM) ---
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 3:16pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/169 "2021-02-17T15:16:57Z")
Volition is not a random act. Volition is the exercise of choice. I am to choose between A and C. I have experienced A and C so I generalize and create B. I hold patents as a result of exercising that process. Of course the network is just a regression mechanism that creates all the points on a control surface from sparse data. It’s not really magic. But it projected data points that it had not experienced and it had the ability to act on them. The algorithm is just a way of explaining how it happened. ---
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 3:36pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/170 "2021-02-17T15:36:49Z")
Interesting video. Yes, it’s just a circuit. The brain is an analog net. The brain is not numerical and it is not binary. It’s output is a gradient. ---
**Author:** ![Babale](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/babale/32/15666_2.png) [@Babale](https://boards.straightdope.com/u/Babale)\ **Post date:** [February 17, 2021, 3:44pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/171 "2021-02-17T15:44:41Z")
Certainly, but it is still deterministic in my book. ---
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 3:48pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/172 "2021-02-17T15:48:42Z")
Perhaps so. Are you saying that if there is a mechanism involved and it acts on data, it’s deterministic? Unless it’s magic or a soul, it’s deterministic? ---
**Author:** ![Babale](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/babale/32/15666_2.png) [@Babale](https://boards.straightdope.com/u/Babale)\ **Post date:** [February 17, 2021, 3:52pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/173 "2021-02-17T15:52:44Z")
No, it could be random. And for example, if it turns out that on the subatomic level true randomness exists, and can propagate up to the macroscopic level and influence things in our everyday world, then randomness can influence our thoughts. I also agree with @begbert2 's point, that in fact even if there was some magical soul making decisions of its own “volition”, if that soul is deciding what to do next based on internally consistent (even if sometimes irrational) processes, that doesn’t allow one to escape determinism. ---
**Author:** ![Max\_S](https://avatars.discourse-cdn.com/v4/letter/m/46a35a/32.png) [@Max\_S](https://boards.straightdope.com/u/Max_S)\ **Post date:** [February 17, 2021, 5:40pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/174 "2021-02-17T17:40:52Z")
> [@begbert2](#): > > Which can easily be demonstrated by the fact that all the proponents of it flatly refuse to even consider how it could possibly work enough to accomplish what it supposedly does. The same criticism can be levied against randomness as event causation. The same criticism applies to literally every theory of event causation except infinite regress; “how it works” implies that you are looking for causation. A non-deterministic process is, by definition, beyond the scope of _determining_ how it works. Truly random selection is one example of a non-deterministic process. Agency is another. > [@begbert2](#): > > My first paragraph was attempting to point out that virtually every reason we have for employing a justice system applies equally well to humans, lakes, and murderous killbots. […] If something is committing crimes, it doesn’t matter why they’re doing it […] That’s your opinion, not mine. Inanimate objects like lakes do not have perception, and therefore do not experience pain or pleasure, and therefore cannot experience _eudaimonia_ (religious deities excepted, of course). We can talk about the _telos_ and _aretē_ of objects (for they do have purposes or excellences), but we cannot speak of their loves and desires. Thus objects cannot be moral actors and we cannot assign them moral culpability. Of course it matters why some_one_ commits a crime, especially within the framework of virtue ethics. The _coup de grâce_ comes to mind. Involuntary acts also come to mind - you may recall this example: > > [@Any atheists here who believe in free will?](https://boards.straightdope.com/t/any-atheists-here-who-believe-in-free-will/838240/112): > > > > Let’s say you had a neurological disorder which caused you to constantly pinch and unpinch your thumb and forefinger as if you were using a television remote. This specific disorder also makes it so that your brain is essentially hotwired to consciously perform this action; you _want_ to pinch your fingers together. I place an unarmed button-style nuclear detonator in your hand, such that it cannot be removed, and tell you not to press the button or millions of people will die. Your neurological disorder means you want to press the button anyways, so you do so. Your “decision” takes place entirely within your head. I tell you the button will be armed in one minute. One minute later some city blows up. I am obviously a monster, but I escape while you are caught by the authorities. Assuming you can prove exactly what happened, do you have any sort of culpability? > [@begbert2](#): > > I’ve heard nature get anthropomorphized numerous times People do personify nature, and pass moral judgements upon it, but - ruling out belief in religious deities - _I_ see that more as “if nature was a person, she would be […]”. A lake cannot properly be immoral. To say “the lake is immoral for drowning a man” is to mean “if the lake were a person, she would be immoral because she drowned a man”. (even that is not an airtight accusation of immorality) In the case of religious deities, such as a lake god, moral judgement of the lake would actually be moral judgement of the god. > [@begbert2](#): > > But supposing you’re correct, that does raise the question of what you consider the purpose and function of morality. In _this_ sub-discussion, the purpose of justice and the function of morality was already given: > > [@jcklpe](#): > > > > But these days I adopt a more restorative idea of justice. The purpose of “justice” has changed from “punish evil, reward the good” to “heal the sick”. […] In my view all moral issues are a matter of “disordered loves”, as Augustine put it. Instead of seeing Justice as a matter of punishing evil or rewarding good, I see it as an attempt to bring reality back into alignment with telos. We are dealing with a form of virtue ethics where moral judgement is to identify whether one’s loves/desires are in alignment with _telos_ (purpose, nature). The purpose of moral judgement in this context is to “bring reality back into alignment with telos”, as a prelude to _restoration_ or _rehabilitation_. > [@begbert2](#): > > if you think that deterministic agents never consciously weigh factors _I_ don’t think deterministic “agents” are conscious at all. This is a simple deduction given my personal philosophy of consciousness. ~Max ---
**Author:** ![begbert2](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/begbert2/32/3357_2.png) [@begbert2](https://boards.straightdope.com/u/begbert2)\ **Post date:** [February 17, 2021, 5:42pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/175 "2021-02-17T17:42:06Z")
I agree - the operations going on inside a neural net are complicated and difficult/impossible to follow from the outside, but each tiny substep certainly falls into one of two categories - either it is the result of a prior state, or it is derived from randomity. And if we’re talking about the sort of neural net that you’d find on a computer, this fact is _really really_ explicit, because modern computers are inherently deterministic. Not only do they have algorithms which they follow, but the hardware itself is designed to eliminate random noise in the processes it runs. So when something on a computer has randomity in it, it is done very very deliberately and in a very known way: it either comes from external inputs, or from an internal random number generator. (Keeping in mind, of course, that most computer random number generators aren’t random at all.) ---
**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\ **Post date:** [February 17, 2021, 5:42pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/176 "2021-02-17T17:42:53Z")
Surely we must consider the possibility that free will is theoretically possible, but humans are incapable of it, so efforts to model human states will never reveal the evidence you seek 🙂 ---
**Author:** ![Crane](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crane/32/3495_2.png) [@Crane](https://boards.straightdope.com/u/Crane)\ **Post date:** [February 17, 2021, 5:46pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/177 "2021-02-17T17:46:43Z")
Babale, Are you proposing that: _All mechanisms are deterministic._ Your view is aligned with Harari. But Harari takes that position because he desperately wants his homosexuality to be determined rather than a life style choice. He announces that position early in the book. His literary struggle with it is a matter of choice .: free will or at least volition. ---
**Author:** ![Max\_S](https://avatars.discourse-cdn.com/v4/letter/m/46a35a/32.png) [@Max\_S](https://boards.straightdope.com/u/Max_S)\ **Post date:** [February 17, 2021, 5:57pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/178 "2021-02-17T17:57:58Z")
> [@Mijin](#): > > Rather than talk about what things _don’t_ decide free choices, what things _do_ ? You’ve answered your own question. Freedom of choice means the agent is free to decide - no process or mechanism will decide on behalf of the agent. There is no answering the “how” question, just like you can’t answer “how” totally random probability actually works. Hence the third category of event causation: infinite regress, random probability, and free agency. ~Max ---
**Author:** ![Max\_S](https://avatars.discourse-cdn.com/v4/letter/m/46a35a/32.png) [@Max\_S](https://boards.straightdope.com/u/Max_S)\ **Post date:** [February 17, 2021, 6:03pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/179 "2021-02-17T18:03:14Z")
> [@jcklpe](#): > > Virtue is virtue. It value is as an end of itself. Agency doesn’t change that. Neither does moral agency (moral culpability). I may have misunderstood your point - are you saying culpability doesn’t carry moral weight? ~Max ---
**Author:** ![Max\_S](https://avatars.discourse-cdn.com/v4/letter/m/46a35a/32.png) [@Max\_S](https://boards.straightdope.com/u/Max_S)\ **Post date:** [February 17, 2021, 6:06pm UTC](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877/180 "2021-02-17T18:06:09Z")
> [@MrDibble](#): > > > [@Max\_S](#): > > > > FWIW I’d bet that the majority of people, both worldwide and just in the U.S., believe in souls and free will. > > Naked appeal to popularity? I am a little disappointed at that. > [@begbert2](#): > > The majority of people both worldwide and in the US believe in stupid bullshit of one flavor or another. Often multiple different flavors of stupid bullshit simultaneously. > > Come now, at this instant we’re pretty much literally having a discussion about a specific arcane branch of theology. You can’t possibly be surprised if the argument “lots of people believe in it” doesn’t fly with everyone present. That line was a response to this, not an argument in support of my position: > [@begbert2](#): > > Bolding mine - I haven’t generally heard libertarianists _admit_ that they are pinning their free will argument on supernatural souls, because that’s a great way to get scoffed at. ~Max [Previous page](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877.md?page=8) [Next page](https://boards.straightdope.com/t/free-will-with-a-little-bit-of-science-tm-c-r/931877.md?page=10)